A Time Machine to Understand Ocean Health

The white-tailed tropic bird was one of eight species from the North Pacific included in the Ocean Memory Lab study. Photo courtesy U.S. Fish and Wildlife Service.

For scientists seeking to understand how the ocean is changing, perhaps the ideal research instrument would be a time machine. Absent such technology, the Monterey Bay Aquarium has been working to create the next best thing. It’s a new facility called the Ocean Memory Laboratory.

For the lab’s inaugural project, researchers have put together a dataset of the feeding habits of eight species of seabirds over the span of almost 130 years. They analyzed archived feathers dating as far back as 1890, using a technique called compound-specific stable isotope analysis, to better understand how the birds’ diets shifted in response to factors ranging from competition with humans to the changing climate.

“In the grand scheme of things, in our field of science, even 10 years of data is encouraging,” says Tyler Gagne, an assistant research scientist at the Aquarium and lead author of the new study, published February 14 in Science Advances. “This is a 130-year-long dataset, which is really amazing.”

Data, data everywhere

The study exemplifies the promise of the Ocean Memory Lab—the brainchild of Aquarium science director Dr. Kyle Van Houtan, who co-authored the publication together with two colleagues based in Hawaii, Dr. David Hyrenbach of Hawaii Pacific University and Molly E. Hagemann of the Bishop Museum in Honolulu.

Identifying novel sources of long-term data is at the heart of the lab’s mission, Kyle says, because conservation projects often lack an informed baseline of ecosystem health to compare against.

“What are the conservation targets? What are we managing for? How do we know when we’re done?” he asks. “We often don’t have enough data or a sufficiently long-term record to provide informed answers to those questions.”

The solution, as Kyle sees it, may lie within the creatures themselves—or more precisely, in the chemistry of their tissues, which can record what they were eating, as well as clues about the surrounding ocean.

For Kyle, it’s a lot like having access to time-traveling drones equipped with sensors to gather data about past ocean conditions. Instead of building robotic drones and electronic sensors to collect information, Kyle and his research team rely on the record left behind in the tissues of long-dead marine creatures.

“Our task is a sort of reverse engineering—to find the sensor within the plant or animal, see what it’s recording, and then translate that into useful information,” he says.

The seabird feather paper is one example. From preserved fronds of kelp to otter teeth, shark vertebrae and whale earwax, Kyle says opportunities to glean details about the ocean’s past are abundant.

“So much data already exists,” he says. “There’s high quality data literally all around us. There are so many opportunities—you just have to pay attention.”

Analyzing the chemistry of feathers catalogued and stored decades ago in museum collections and other repositories is just one way of putting that data to use, he says.

“My hope is that we are not only taking advantage of the work of our predecessors and essentially standing on the shoulders of giants, but also giving new scientific value to this wealth of data that have been archived for such a long time,” Kyle says.

Old feathers and new analytical tools

The feather analysis is a perfect illustration.

It hasn’t always been easy to get reliable data on what seabirds eat, since they travel vast distances and spend much of their time far from shore. Researchers often use data from fisheries observers, but those data comes with inherent biases, Tyler says.

“Rather than just randomly sample from an ecosystem, a fisheries harvest is based on social and market forces,” he explains. “What fish are we most interested in buying and eating? What offers the best return on investment from fishing? While birds of course have some inherent bias in their foraging, they don’t have the bias of human commercial fisheries.”

The new study looked at eight species of Pacific seabirds, from the Laysan albatross to Bulwer’s petrel.

“These eight species eat a range of diets and forage at different distances from land,” Tyler says. “Some birds are relatively heavy and stay somewhat close to where they nest. Others can fly quite efficiently, search a much larger area of the ocean, and they may tell a story about what is happening farther away.”

With each snapshot of a bird’s diet comes a glimpse into the historic past of the ocean’s health: How abundant were different species of fish at different times? And to get a sense of what the owner of a particular feather was eating, you need only a tiny sample: “milligrams— just a fraction of your pinky fingernail,” Tyler says.

By grinding that feather sample into a powder and studying it using stable isotope analysis, researchers can tease out information about a bird’s diet from tracer particles within the feather. For instance, nitrogen is accumulated in a tiny phytoplankton that gets eaten by zooplankton, which is then eaten in turn by a sardine, which might then get eaten by another fish, and ultimately by a seabird.

“Nitrogen isotopes have a way of enriching that corresponds to an organism’s position in the food web,” Tyler explains. “By looking at the nitrogen isotopes in each amino acid (the building blocks of proteins) we can calculate an animal’s exact position in the food web.”

Parsing the ensuing data from a large dataset (134 birds in total) is difficult. That’s where machine learning came in handy—utilizing algorithms to find patterns in data that older analysis tools perhaps would miss.

The computer analysis uncovered some intriguing patterns.

Declining trophic level, increasing data

The eight North Pacific species in the study tended, on average, to decline in mean trophic level, meaning they are eating lower on the food web today than they did in the past. In practical terms, that means birds that would have found and eaten fish a century ago are now more likely to find only squid. (Because of their short lifespans and dependence on environmental factors, squid populations can vary highly from one year to the next.)

In line with other research, the study ultimately argues that fishing pressure by commercial fleets is a key factor, along with climate, in driving seabirds to lower positions in the food web. That the Ocean Memory Lab was able to scientifically demonstrate a novel way to assess the scope of this problem is significant, Tyler says.

“It certainly offers deeper insight,” he says. “What do seabirds tell us that maybe the fishery-dependent metrics didn’t? It seems quite a bit in this case.”

As the environment continues to change, data that provide context and clarity to conservation efforts will be invaluable. In its drive to uncover such information, Aquarium researchers are just getting started at the Ocean Memory Lab.

By Daniel Potter, Monterey Bay Aquarium

latest issue

From groundbreaking marine instrumentation to metocean data visualization and lab-to-market efforts carried out by research-industry partnerships, the ongoing expansion of…

Search